The most personal data you will ever generate should never leave your home.

No. We produce your raw genomic data for your own learning and use. It is not a validated clinical test, and you should not make medical decisions from it. If something concerns you, take it to a clinician who can order accredited diagnostic testing.

Those services use a genotyping microarray, a chip that checks a pre-selected set of roughly 600,000 to 1 million spots in your DNA, well under 0.1% of your genome and only positions chosen in advance. Whole-genome sequencing reads essentially all of it, about 3 billion base pairs, and captures variation a chip cannot see: rare and novel variants, insertions and deletions, large structural changes, and mitochondrial DNA. A chip looks up answers to questions someone decided to ask in advance; sequencing reads the whole book.

Your genome is the raw data behind a wide range of things people explore: how you might respond to certain medications (pharmacogenomics), recessive conditions you carry that can matter for family planning, fine-grained ancestry, and traits and predispositions studied across thousands of genes. Because you hold the complete data, you can revisit it as the science improves, applying new discoveries to a genome you sequenced years earlier. We provide the raw data for your own learning, not a diagnosis, so take anything health-related to a clinician.

Plenty. Most common traits and diseases come from many genes acting together with environment and chance, so a genome shows predispositions and probabilities, not certainties. It will not tell you whether or when you will develop a condition, it cannot account for non-genetic factors like diet, exercise, and exposures, and a few regions of the genome are still genuinely hard to read. Treat it as one rich input about yourself, not a verdict.

Sequencing reads each part of your genome many times over, not once. 30× means every position is read about 30 times on average. That redundancy is what makes the result trustworthy: overlapping reads cancel out random errors and let us call your variants with confidence. 30× is the depth large-scale and clinical-grade genomics treat as the standard for a reliable whole genome.

We use Oxford Nanopore technology, which reads long, continuous stretches of your native DNA instead of stitching together tiny fragments. Long reads resolve repetitive and complex regions, reveal large structural changes that short-read methods miss, allow variants to be phased (knowing which changes sit on the same copy of a chromosome), and read DNA methylation directly, all in a single run on your unmodified DNA.

Your genome is read to 30× average coverage and basecalled with Oxford Nanopore's latest models, where modern reads reach well above 99% accuracy, then aligned to the human reference. Reading every position about 30 times is what lets us call your variants with confidence, and 30× is the standard depth used in large-scale genomics for a reliable whole genome.

Five steps, all in your home. First, a painless cheek swab. Second, on-site library prep: we extract your DNA, check its quality, and attach the adapters the sequencer needs. Third, sequencing: your DNA strands pass through nanoscopic pores in the PromethION 2 Solo, which reads each one as a tiny electrical signal over roughly 48 to 72 hours. Fourth, basecalling: our self-contained NVIDIA GPU box runs Oxford Nanopore's Dorado software to turn that signal into your A, C, G and T sequence in real time, then aligns it to the human reference. Fifth, handover: we confirm coverage and quality, write your genome to a drive you keep, and take every instrument with us.

You receive your raw, aligned genome as standard BAM and FASTQ files on a drive you own. That is the most valuable and future-proof form of the data: you can run it through the analysis tools and services of your choice and revisit it as the science advances. We deliberately do not lock you into a single interpretation.

Nothing your genome touches is ever online. The sequencer and the NVIDIA compute box are air-gapped, with no network connection during your session, so nothing can be uploaded, intercepted, or synced. Basecalling and alignment, the steps that usually happen in a cloud, run entirely on the box in your living room. Your finished genome is written only to a drive you keep, and we leave with the hardware holding no copy. We are not promising to delete your data later; we are built so a copy never leaves the room in the first place.

A trained medical professional, in your home, from the first swab to the final handover.

A simple cheek swab. Painless, and enough for a complete genome.

After you reserve we reach out to schedule. The visit is one setup, roughly 48 to 72 hours of sequencing in your home, and a short return visit to hand over your data.

Yes. The $500 deposit is fully refundable up to one week before your scheduled sequencing date. For pre-orders, that cutoff is July 7th. Email me@pandeymikhil.com and see our refund and cancellation policies.

San Francisco and New York City. Because we come to you in person, we operate city by city and are expanding to more over time.